No guru. No hype. Just a techie running the actual 2026 agent stack — self-hosted agents, background PRs, parallel builds — and showing every win, failure, and fix on camera.
Every build is real, recorded, and ships a working artifact.
60–90 second reels: the problem, the agents working, the result. The hook is always a real artifact.
Full-length sessions on YouTube. Unpolished, thinking out loud — the setup, the break, the fix.
Every working .cursorrules, SKILL.md, and agent config lands in the public repo. Take them, run them.
The tools most AI accounts only talk about — I run them.
One email per build series — what I built, what broke, and the configs to run it yourself. No noise.
Build systems. Evolve the operator.
Better. Faster. Smarter. anytime, anywhere, on any device.